Basque butterflies again

[Larry Trask]

This is a joint response to the comments made by Ed Selleslagh and by
Lloyd Anderson.

First, I endorse Ed's observation that words for 'butterfly', at least
in European languages, often appear to exhibit a moderately high degree
of instability, rendering them of questionable value in investigating
linguistic prehistory.

Second, and far more importantly, I am not happy with attempts at
considering the numerous Basque words for 'butterfly' in isolation
from the rest of the Basque lexicon.

The 'butterfly' words in Basque, or most of them, exhibit a number of
phonological and distributional characteristics which are entirely
typical of expressive formations but not at all typical of ordinary
lexical items: unusual length, unusual segments and sequences, severe
localization, great and unusual variation in form, and other things.

Take the mainly Lapurdian forms <pinpirin(a)> ~ <pinpilinpauxa>.
Lapurdian, almost alone among the varieties of Basque, is very fond of
expressive formations in <pinp->, <panp-> and <punp->.  Here are some
examples of such expressive forms:

	<pinpa> 'bounce'
	<pinper> 'back'
	<pinpili-panpala> 'favorite'
	<pinpin> ~ <pinpirin> 'bud' (of a flower)
	<pinpin(t)xori> 'tit' (bird)
	<pinpiñeka> ~ <pinpirin> 'garfish'
	<pinpi-panpa> 'bang!'
	<pinpoil> 'somersault'
	<panp> 'sound of falling'
	<panpa> 'rag doll'
	<panpa> 'thud', 'heartbeat'
	<panpala> ~ <palpara> 'chatterbox', 'long-winded person'
	<panpalina> 'a kind of bell'
	<panpan> 'hitting' (nursery word)
	<panparriko> '(a) blow'
	<panpina> ~ <panpa> 'doll'
	<panpotsa> ~ <panpotza> 'heartbeat', 'elegantly dressed'
	<panpoxta> ~ <panpotta> 'petite, plump and pretty' (of a woman)
	<panpun> ~ <papun> 'very large mushroom'
	<punp> 'sound of a fall', 'sound of a gunshot'
	<punpa> '(a) fall', 'peashooter'
	<punpa> 'in a slow, trudging manner'
	<punpa> 'large bell'
	<punpa> ~ <punpe> 'bounce' (of a ball)
	<punpilitaka> 'hammering'
	<punpoillo> 'tuft', 'tassel'
	<punpula> ~ <punpulla> ~ <punpullo> '(a single) tear', 'dewdrop',
		'bubble', 'bulb'
	<punpullka> 'somersaulting'
	<punpullu> 'clog' (footwear) (nursery word)
	<punpun> 'poppy'

These words have no identifiable morphological structure, and they contain
no recognizable morphemes, apart from the adverbial <-ka> found in a couple
of them.  They are nothing but sequences of speech sounds.

Now, I submit that the Lapurdian butterfly words in <pinp-> belong strongly
to this group of expressive formations, and that their origins should
therefore be sought in this specifically Lapurdian pattern of coining
expressives, and not among vague resemblances in implausibly far-flung
languages.

It appears that the Lapurdians, some centuries ago, settled on an
expressive pattern consisting of <pinp->, <panp-> or <punp-> followed
by more or less arbitrary sequences of agreeable sound in order to coin
expressives.  Items of such a form are all but unknown in other varieties
of Basque.  And, of course, native Basque lexical items with the slightest
claim to antiquity never begin with /p/ (the ancient Basques apparently
couldn't pronounce initial /p-/) and never contain the sequence /np/ (which
was categorically voiced to /nb/ in Lapurdian and in most other dialects in
the medieval period).

A few further comments.  The word <bitxi> 'pretty', 'pretty little thing',
'ornament', 'jewel', apparently present in the large group represented by
<bitxilota>, is of uncertain origin but is often suspected of being a
palatalized form of <bizi> 'alive', 'living'.  The suggested semantic
development is 'living' > 'moving' > 'flashing', 'sparkling', > 'pretty',
'jewel', or something similar.

So, as Ed suggests, <bizi> is very likely the *ultimate* source of the first
element, but it cannot be the *direct* source.

The final sequences <-tola>, <-dola> and <-tera>, found in some of the words,
are described by Ed as familiar derivational or diminutive suffixes.  Not in
Basque, I'm afraid.  No such word-forming suffixes as these are attested in
Basque in any function at all: they appear to be no more than arbitrary
extensions.	

Finaly, I am amused by Ed's suggestion that the <zintz-> of <zintzitoil> (and
variants) might represent "the noise of a flying insect".  Butterflies don't
make any noise.

I therefore conclude that Lloyd's efforts at linking some of the Basque words
to words in other languages (and also to one another) are without foundation.
The Basque 'butterfly' words are numerous and severely localized; they conform
strongly to observed patterns for coining expressive formations; and they can
scarcely be of any antiquity.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk